CVS viewer breaks binary files

It’s not possible to download non-textual data from CVS (for example, these), because the viewer attempts to parse them as Unicode (even if the content-type field in the URL is set to application/octet-strem), mangling them beyond all usefulness.

Any chance this could be fixed?

Not so much the CVS viewer’s fault I think, as the fact that the files weren’t marked as binary when they were imported.

Rob: when you’re next working on that component, can you run the “cvsbinaries” perl script on that component’s sandbox please?